Quanta Cloud Technology (QCT), a global provider of hyperscale cloud and hyperconverged infrastructure, has launched a new series of bare metal switches (BMS) supporting 25Gbps and 100Gbps – for hyperscale data center networking solutions.
These new switches – the QuantaMesh T7032-IX1 and the QuantaMesh T4048-IX2 – are powered by Broadcom Tomahawk chips and would deliver a 25 percent performance improvement compared to 10/40G switches.
Cloud infrastructure operators are evolving from 10G and 40G switches to 25G and 100G hardware, and these new BMS offerings from QCT address and support data center customers’ needs to optimally pair networking software with the hardware of their choice.
The bare metal Ethernet switches would offer not only low cost of ownership (TCO), but also the ability to optimize software for best performance. By deploying the QuantaMesh 25/100G Ethernet switch, data center operators can upgrade to 25G – for top-of-rack (ToR) configurations – and 100G at the middle-of-row (MoR) and end-of-row (EoR) configurations based on spine-and-leaf infrastructure for growing server/storage workloads.
QCT’s QuantaMesh Ethernet switches feature integrated baseboard management controller (BMC), which usually resides in servers. This would allow administrators to use the same management tool and protocol for management of all data center assets, enabling further savings on operational costs.
The QuantaMesh T7032-IX1 supports 32 QSFP28 (10/40 or 25/50/100GbE speed) ports in a 1U size to deliver ToR or spine switch performance. The QuantaMesh T4048-IX2 supports 48 SFP28 (10/25GbE speed) downlink ports and 8 QSFP28 (10/40 or 25/50/100GbE speeds) uplink ports in a 1U size in ToR switch position to perform 1.5:1 oversub performance.
QuantaMesh BMS Ethernet switches are based upon open hardware designs with ONIE (Open Network Install Environment) pre-loaded to support multiple NOS. Cumulus Linux OS and ICOS NOS are certified to install for different networking applications. These networking products would be ideal for applications that require high performance, high availability, fast scale out, low latency performance and continuous serviceability.